Action does not enhance but attenuates predicted touch

Abstract

Dominant motor control theories propose that the brain uses efferent information to predict and attenuate the somatosensory consequences of actions, referred to as sensory attenuation. Support for this model comes from psychophysical and neuroimaging studies showing that touch applied on a passive hand elicits attenuated perceptual and neural responses if it is generated by actively tapping with one’s other hand, compared to identical touch from an external origin. However, recent experimental findings have challenged this view by providing psychophysical evidence that the perceived intensity of touch on the passive hand is enhanced if the active hand does not receive simultaneous tactile stimulation with the passive hand (somatosensory enhancement) and by further attributing attenuation effects to the double tactile stimulation of the hands upon contact. Here, we directly contrasted the hypotheses of the attenuation and enhancement models regarding how action influences somatosensory perception by manipulating whether the active hand contacts the passive hand. In three preregistered experiments, we demonstrate that action does not enhance the perceived intensity of touch (Experiment 1), that the previously reported “enhancement” effects are driven by the baseline condition used (Experiment 2), and that self-generated touch is robustly attenuated regardless of whether the two hands make contact (Experiment 3). Our results provide conclusive evidence that action does not enhance but attenuates predicted touch. These findings prompt a reappraisal of recent experimental findings upon which theoretical frameworks proposing a perceptual enhancement by action prediction are based.